particular, the results support the proposition that behavioural factors are important and
significant determinants of whether or not an individual experiences financial distress.
Financial Literacy: Finally, in the sixth column, I focus only on UK respondents and
assess the impact of being financially literate on the incidence of financial distress. Con-
trolling for all the demographic, income and behavioural factors already discussed, I find
that financially literate individuals in the sample are less likely than financially illiterate
individuals to experience financial distress. The results suggest that the greater the number
of questions answered correctly by respondents, the lower the probability that respondents
will have experienced financial distress. For example, the probability of getting into fi-
nancial difficulties is about 8 per cent lower for people who get five of the six questions
on financial literacy correct, relative to people answering three or less questions correct,
and this result is statistically significant.7 It is worth noting that the behavioural factors
remain significant in this regression while the size of the coefficients on two of the variables,
impulsiveness and time preference, are larger than those on the financial literacy variables.
8
Summary: The analysis so far points to a number of key results:
1. Firstly, demographic and economic factors matter for financial distress. The effects
are as expected, and in line with the studies surveyed in Section 2. In particular,
relationship breakdown, having dependent children, being unemployed and having
outstanding debt all increase a person’s probability of getting into financial difficulties,
while a college education and higher income reduce the probability. The probability
of financial distress also increases with age, but only up until the late-30s, after which
point the probability falls.
7 I also run a regression where financial literacy is instead captured by dummy variables for each of the
six questions (where the dummy variable equals one if the respondent gets the question correct and zero
otherwise). The results, which are available from the author, show that an ability to distil information
from investment related graphs is most important (in terms of statistical significance) for whether or not a
respondent gets into financial difficulties.
8 Furthermore, I run a probit regression of financial distress on the financial literacy variables and the
remaining demographic and socio-economic variables (excluding the behavioural variables). The absolute
size of the marginal effects on the financial literacy variables are higher by about 2 percentage points. This
shows that a failure to account for behavioural factors over-estimates the importance of financial literacy
on the probability of getting into financial trouble.
20